Effects of Pressure, Temperature, Fluid-Rock Interactions, and
Phase Changes on the Physical Properties of Geothermal Reservoir
Rocks: the Experimental PerspectiveProceedings World Geothermal
Congress 2010 Bali, Indonesia, 25-29 April 2010
1
Effects of Pressure, Temperature, Fluid-Rock Interactions, and
Phase Changes on the Physical Properties of Geothermal Reservoir
Rocks: the Experimental Perspective
Harald Milsch1, Erik Spangenberg1, Siegfried Raab1, Ansgar
Schepers1, Guido Blöcher1, David Bruhn1, Líney H. Kristinsdóttir2,
Ólafur G. Flóvenz2 and Ernst Huenges1
1GFZ German Research Centre for Geosciences, Telegrafenberg, 14473
Potsdam, Germany 2Icelandic GeoSurvey (ÍSOR), Grensásvegi 9, 108
Reykjavík, Iceland
[email protected]
ABSTRACT
Within the GFZ, Section “Reservoir Technologies”, there is a
long-standing tradition in experimental rock physics. We maintain
several high pressure (100 MPa) and high temperature (200°C)
flow-through apparatuses to address questions arising in geothermal
reservoir characterization, evolution and sustainability. Various
research programs are conducted both site specific and process
oriented. So far, the types of fluid-rock combinations explored,
relate to the geothermal wells in Groß Schönebeck (Germany),
Hengill and Krafla (Iceland), and Anqua and Radicondoli (Italy),
respectively. With regard to particular parameters and processes
related to these reservoirs we experimentally investigated:
1) The pressure dependence of rock transport properties and their
interrelations (Germany), where three different models relating
permeability and electrical conductivity were tested and the
appropriateness of an individual model showed to be rock-type
dependent. 2) The effect of dissolution-precipitation reactions on
rock permeability (Germany), where in a series of long-term
flow-through experiments at various chemical fluid compositions
even after six months of flow no significant change in permeability
in neither direction was observed. 3) The temperature dependence of
electrical rock conductivity and seismic wave velocities (Iceland
and Italy), where the values derived for the temperature
coefficient α, were in the range 0.027-0.160 1/°C indicating the
predominance of interface conduction regardless of the respective
alteration stage. P-wave velocities systematically decreased with
temperature and were in the range of 4.4 km/s (25°C) and 3.4 km/s
(250°C). 4) The evolution of electrical rock conductivity in a
fluid- rock disequilibrium (Germany and Iceland), where observed
transient conductivity changes are interpreted as alterations of
the fluid-rock interface properties and/or of pore fluid
compositions resulting from dissolution processes. 5) The
petrophysical signature of a water-steam phase transition within
the pore space (Germany and Iceland), where it showed that the
conduction mechanism (fluid vs. interface conduction) controls the
pattern of electrical conductivity variations as steam saturation
changes.
Here, we provide an outline of the technical features of our
apparatuses, background information on the particular geothermal
setting as well as experimental details and results of the
individual projects outlined above.
1. INTRODUCTION
Effective energy production from geothermal reservoirs requires the
physical properties of the host rock to be characterized as
precisely as possible. This can be achieved by an appropriate
combination of production and/or injection tests and
borehole-logging techniques. Rock physical experiments, in
addition, provide a valuable complementary method to investigate
particular processes associated with mechanical and thermodynamic
changes induced during operation. Not least, the results of such
investigations can be included in hydro-thermo-mechanical- chemical
(HTMC) simulation codes to derive statements on reservoir
productivity, sustainability, and best-practice operation.
For hydrothermal settings the key parameter for reservoir
productivity and indirectly its thermal evolution is the matrix
permeability which in turn is affected by changes in effective
stress through changes in pore space geometry (e.g. Jaeger and
Cook, 1976; Blöcher et al., 2009). Furthermore, production-induced
thermodynamic fluid-rock disequilibria can severely damage
permeability when mineral precipitates or other fines plug the pore
throats (e.g. Civan, 2000).
Electrical rock conductivity, as a second important rock transport
property, is widely used in geophysical borehole- prospecting to
indirectly obtain information about the formation permeability
given an appropriate correlation function (e.g. Martys and
Garboczi, 1992). In contrast to permeability, electrical
conductivity is strongly dependent on temperature and the
predominant individual conduction mechanism which can be surface or
fluid dominated (Ruffet et al., 1995). Differences in the
conduction mechanism yield different temperature coefficients (e.g.
Revil et al., 1998) which, in addition to the pore fluid chemistry,
have to be known to adequately determine the formation factor
(Archie, 1942). In volcanic environments such distinctions can also
assist to correctly map resistivity variations associated with
different stages of alteration (Flóvenz et al., 1985).
Finally, in a thermodynamic disequilibrium the electrical rock
conductivity can be altered by either changes in pore fluid
chemistry (Piwinskii and Weed, 1976) and/or phase transitions (e.g.
water to steam) (Roberts et al., 2001). In both cases electrical
conductivity will be time dependent, in the second case it will
also depend on the residual saturation provided that conduction is
fluid dominated.
2. EXPERIMENTAL METHODOLOGY
For process-oriented studies in the geothermal context testing
under realistic in situ pressure and temperature conditions is
necessary. This also includes the selection of
Milsch et al.
2
an appropriate sample material and the use of the respective
formation fluids as well as a continuous fluid flow through the
specimen. In addition, as an evolution of the rock transport
properties depends on the rates of the transformations involved,
stable experimental conditions have, eventually, to be maintained
over longer time periods. Finally, it is desirable not only to
measure changes in the petrophysical transport properties (e.g.
permeability) but one should also be able – for explanatory
purposes – to complement these investigations with pore fluid
sampling and an analysis of the fluid chemistry. To meet these
requirements two identical devices have been set up at the GFZ
being a refined derivative of an older concept described by
Kulenkampff et al. (2005). Continuous flow experiments, so far,
have been performed over a maximum of six months.
The apparatus consists of an internally heated oil-medium pressure
vessel and a connected pore fluid system. The sample assembly is
inserted vertically into the vessel. Both confining- and pore
pressure are generated with piston- cylinder type syringe pumps.
The apparatus allows simultaneous and continuous measurements of
permeability, electrical conductivity as well as P- and S- wave
velocities at maximum confining pressures, pore pressures and
temperatures of 100 MPa, 50 MPa and 200 °C, respectively.
Typically, the sample size is 30 mm in diameter and 40 mm in
length.
The permeability of a rock is determined by a steady state method
making direct use of Darcy’s Law (e.g. Darcy, 1856; Scheidegger,
1974; Bear, 1988). The electrical conductivity is measured with a
four-electrode arrangement. Silver rings painted onto the samples
at a distance of 25 mm are used as the potential electrodes.
Ultrasonic measurements are performed with piezoelectric ceramics
for both compressional (P) and shear (S) waves, respectively. For
the electrical conductivity- and ultrasonic measurements voltage
signals are impressed with a function generator (Agilent 33220A).
Typically, the voltage is an AC-sine 1.0 V peak-to-peak signal at a
frequency of 13 Hz and a rectangular voltage single burst at 1.0 V
and 400 kHz for conductivity- and ultrasonic measurements,
respectively. For electrical conductivity the input impedance is 10
MOhms. The temperature is measured with two PT-100 sensors, one
close to the top and one close to the bottom of the specimen,
respectively. Generally, one notices a temperature gradient of
approximately 1 - 2°C along the sample, the top side being the
hotter part. The use of corrosion resistant parts throughout the
pore pressure system allows investigations with highly saline
formation fluids which can also be sampled under pressure for
further chemical analysis. Figures 1a, 1b, and 1c show the general
set-up of the apparatus, the individual parts of the specimen
assembly, and the mounted specimen assembly, respectively.
A detailed description of the apparatus and specific measurement
procedures can be found in Milsch et al. (2008a).
3. GEOTHERMAL SETTINGS ADDRESSED
3.1 North German Basin
The technical feasibility of geothermal power production from
hydrothermal (sedimentary) reservoirs with normal geothermal
gradients will be demonstrated by means of the geothermal research
wells Groß Schönebeck (40 km north of Berlin, Germany) using a
borehole doublet as shown in Figure 2.
Figure 1: (a, top) General set-up of the apparatus: (1) and (5)
down- and upstream pore fluid pumps, respectively; (2) reservoirs
for fluid sampling; (3) pressure vessel with internal heater and
specimen assembly; (4) confining pressure pump; (6) fluid
reservoir. (b, center) Parts of the specimen assembly. Silver rings
on the sample serve as potential electrodes for electrical
conductivity measurements. (c, bottom) Detail of the mounted
specimen assembly
The first well Groß Schönebeck EGrSk3/90 was tested to investigate
scenarios of enhancing productivity of thermal fluid recovery from
the underground. In order to complete the doublet system a second
well GtGrSk4/05 with a total depth of -4198 m has been finished in
2007, followed by three stimulation treatments to enhance
productivity (enhanced geothermal system; EGS). For the development
of an optimized effective pay zone the new well is inclined in the
reservoir section by 48° and was drilled in the direction of the
minimum horizontal stress (Sh=288° azimuth) for optimum hydraulic
fracture alignment in relation to the stimulated pre-existing well
EGrSk3/90. Hence, the orientation of the fractures will be 18°
azimuth in the direction of the maximum horizontal stress.
Milsch et al.
3
Figure 2: Location of the research drill site (top) and 3D view of
the two research wells and the geological horizons; the reservoir
is situated in the Lower Permian within a depth of -3850 and -4258
m (bottom)
The reservoir is located in -3850 to -4258 m depth within the Lower
Permian of the North East German Basin. The reservoir rocks are
classified into two rock units from base to top: volcanic rocks
(Lower Rotliegend of the Lower Permian) and siliciclastics (Upper
Rotliegend of the Lower Permian) ranging from conglomerates to fine
grained sand-, silt- and mudstones.
The fault pattern analysis of a 3D structural model indicates
normal to strike slip faulting for the Lower Permian sediments. The
formation pore pressure (pp) is 43.8 MPa, determined by p-T logs at
stationary conditions of the geothermal target horizon (Legarth et
al., 2005). According to the stress relation of normal faulting the
effective mean stress (σmeff) was calculated as 42.9 MPa (Blöcher
et al., 2008). The temperature at this location follows a normal
geothermal gradient and is approximately 150°C. The pore fluid
within the formation is of Ca-Na-Cl type with a high salinity (TDS
≈ 265 g/L). Furthermore and in addition to other species, it has a
non-negligible degree of mineralization with respect to Fe, Ba and
SO4 with concentrations typically around 67.6 mg/L, 25.5 mg/L, and
51.0 mg/L, respectively.
A detailed description of the geology at this site and the
stimulation treatments can be found in Moeck et al. (2008) and
Zimmermann et al. (2008), respectively.
3.2 Iceland
High temperature geothermal fields in Iceland, like in other
volcanic areas of the world show a particular resistivity
structure. Here, a broad resistivity anomaly exists with an
up-doming conductive cap covering a resistive core (e.g. Flovenz et
al., 1985). The conductive cap consists of volcanic rock with
considerable content of conductive alteration minerals, like
smectite, that have high cation exchange capacity. The resistivity
of the cap decreases with increasing temperature until a core of
high resistivity is reached. According to well logging, the
transition from the low resistivity cap to the high resistivity
core coincides with a change in the mineral alteration, i.e. from
smectite to mixed layer clays and chlorite.
Rock samples tested in the present studies originated from three
different geothermal fields and five different wells: (1) the
Ölkelduháls-field (well ÖJ-1), (2) the Nesjavellir-
field (well NJ-17), both in the Hengill volcanic complex near
Reykjavik, and (3) the Krafla-field (wells K-2, KH-1, and KH-5) in
NE-Iceland.
With decreasing depth the samples display different stages of
alteration from the chlorite over the mixed layer clay to the
smectite zone. Rock types are hyaloclastites and basalts. All
fields are freshwater liquid-dominated with fluids of Na-K-Cl-SO4
type and low salinity (TDS ≈ 1 g/L).
4. INDIVIDUAL PARAMETERS AND PROCESSES INVESTIGATED
4.1 Pressure Dependence of Rock Transport Properties and Their
Interrelations
4.1.1 Motivation
Both permeability (k) and (specific) electrical conductivity (σ)
are measured as bulk properties but are in fact defined by the
individual pore structure of a rock. This is evident for
permeability but is also true for the electrical conductivity as
long as the latter is governed by the conductivity of a fluid
within the pore space. Furthermore, the pore structure is affected
by changes in the state of stress acting on the rock. Both
transport properties are therefore dependent on effective stress
or, in the lithostatic (isotropic) case, on effective pressure
(peff). Compared to permeability, electrical rock conductivity is
significantly easier to measure both in the lab and in situ. Thus,
not least for practical reasons, it is desirable to establish a
link between both transport properties and to couple both
parameters through microstructure-related length scales.
Such approaches have been made repeatedly during the last five
decades. Based on geometrical (equivalent channel) models (Wyllie
and Rose, 1950; Paterson, 1983; Walsh and Brace, 1984) as well as
statistical and percolation concepts (Katz and Thompson, 1986,
1987; Guéguen and Dienes, 1989) a relationship (Eq. 1) is obtained
that links both transport properties, where the electrical
conductivity is expressed in terms of the formation factor
(F):
F L ck
12= , (1)
where c and L denote a shape factor and a characteristic length
scale, respectively. The formation factor F here is defined as the
ratio between the electrical conductivity of the fluid (σfl) at the
respective experimental temperature and the measured conductivity
of the rock (σ). The physical meaning of both parameters c and L is
dependent on the respective model and can vary significantly.
The purpose of this study was to test the predictions of Eq. 1
against original experimental and microstructural data obtained for
three different types of sandstone. More specifically, we tested
the models proposed by Walsh and Brace (1984), Guéguen and Dienes
(1989), and Katz and Thompson (1986, 1987) with shape factors (c)
and length scales (L) listed in Table 1.
Furthermore, we compared Eq. 1 with an established empirical
relationship between k and F (Eq. 2; e.g. Brace, 1977; Walsh and
Brace, 1984 and references cited therein) that has been obtained
from investigations on the pressure dependence of the coupled
transport properties:
rF L ck E
4
Table 1. Parameters of the scaling models investigated. c: shape
factor; L: length scale; m: hydraulic radius; Vp: (total) pore
volume; Ap: (total) inner pore surface; rA: average tube radius;
wA: average crack half aperture; lc: length scale defined by Katz
and Thompson (1986, 1987).
where r denotes an empirical, rock dependent parameter following
the notation in Walsh and Brace (1984). The subscript E has been
introduced to distinguish between both length scales defined within
the respective relationship.
4.1.2 Experimental Procedure
For the experiments three different types of sandstone samples were
chosen: (1) Fontainebleau sandstone, a pure quartz arenite, (2)
Flechtinger sandstone, a Lower Permian (Rotliegend) sedimentary
rock quarried from an outcrop near Flechtingen, Germany, and (3)
Eberswalder sandstone, a Lower Permian (Rotliegend) rock cored
during drilling of a prospective gas well (Eb2/76) at Eberswalde,
Germany. The two Rotliegend samples were chosen for their
mineralogical and morphological similarity with the reservoir rocks
at the Groß Schönebeck site where cores were no longer
available.
To enable electrical conductivity measurements 0.1 molar
NaCl-solution was used as the pore fluid. The experiments were
performed in the HPT-permeameter described above. To avoid
disturbance of the measurement by room temperature fluctuations the
experimental temperature was maintained at 40 (± 1) °C.
Both transport properties were measured simultaneously. To
investigate the relative changes of both parameters effective
pressure ramping was performed by successively increasing and
decreasing both confining and pore pressure. The pressures were
varied from 10 to 50 MPa (pc) and 5 to 45 MPa (pp), respectively.
Three full cycles and thus 12 individual ramps were conducted for
each sample. In contrast to electrical conductivity which can be
determined continuously during pressure ramping, permeability
measurements have to be performed stepwise. Depending on the sample
permeability 5 to 10 measurements have been taken along each ramp
at effective pressure intervals ranging from 2.5 to 15 MPa.
Table 2. Sample properties at starting conditions: confining
pressure = 10 MPa; pore pressure = 5 MPa; Temperature = 40°C. φs:
sample porosity measured by saturation at zero effective pressure;
k0: permeability; σ0: electrical conductivity; F0: formation
factor.
Subsequent to each test thin sections were prepared. Two on the
opposing faces of the specimen and one along the core. The sections
were saturated with blue epoxy to allow 2D image analysis to be
performed on binary images (Zeiss Axioplan with Axiocam and KSRun).
The image analysis program yields the measured pore radius
distribution as a function of the (total) cumulative porosity as
well as an average pore radius.
In addition, mercury porosimetry (WS2000, Fisons Instruments) was
performed after the experiments on broken parts of the samples
having a volume of approximately 1 to 2 cm3 each. This independent
method also yields a pore radius distribution of the samples as a
function of the (connected) cumulative porosity as well as an
average pore radius. In addition, the specific inner surface
distribution is calculated from the mercury injection curve.
4.1.3 Results and Discussion
Figures 3a and 3b show the measured normalized permeability and
electrical conductivity, respectively as a function of effective
pressure. Absolute values can then be obtained from the respective
figure and Table 2. Both transport properties are sensitive to
changes in effective pressure. The sensitivity increases from the
Fontainebleau over the Flechtinger to the Eberswalder sandstone.
Furthermore, for all sandstones the sensitivity decreases with
increasing effective pressure. The percental changes of both
transport properties are closely related for the Fontainebleau and
the Flechtinger sandstones. In contrast, for the Eberswalder
sandstone the decrease in permeability at lower effective pressures
is significantly more pronounced than the decrease in electrical
conductivity.
Figure 3: (a) Normalized permeability as a function of effective
pressure. (b) Normalized electrical conductivity as a function of
effective pressure. Normalization has been performed with the
starting permeability (k0) and starting conductivity (σ0) in Table
2, respectively
Milsch et al.
5
The average pore radius decreases from the Fontainebleau over the
Flechtinger to the Eberswalder sandstone. In contrast, the specific
inner surface (AHg) of the Fontainebleau sandstone is significantly
smaller than the ones of the two Rotliegend samples. The pore
radius distribution from mercury porosimetry (Figure 4a) indicates
that is due to the narrow distribution and large pore radius of the
Fontainebleau sandstone. This contrasts the distribution of the two
Rotliegend samples implying significant morphological differences.
Their larger specific inner surfaces (Figure 4b) are principally
due to contributions from pore radii between 0.2 – 2.0 µm.
Figure 4: (a) Cumulative porosity and (b) cumulative specific inner
surface as a function of the pore radius measured with mercury
porosimetry
Testing of the scaling models described above has been performed by
permutation of all three parameters measured (permeability,
formation factor, and length scale) calculating one property from
the two others. It showed that none of the tested models could
reproduce any of the experimentally or microstructurally determined
parameters adequately within experimental precision. Additionally,
there was no clear preference for one of the models as for each of
the rocks there was a different model with the closest numerical
agreement. It also showed that models that are based on a tube
geometry yield better results than crack models, which in this
study is solely related to the differences in the shape factor.
Furthermore, the compilation demonstrates that the use of the
average throat radius as the length scale in the model by Guéguen
and Dienes (1989) yields better results than the average pore
radius itself.
The observed inaccuracy of the models by Walsh and Brace (1984) and
Guéguen and Dienes (1989) might be related to the assumption that
both hydraulic and electrical transport follows the same flow paths
(David, 1993; Van Siclen, 2002). This implies transport property
dependent tortuosities (Walsh and Brace, 1984) and percolation
factors
(Guéguen and Dienes, 1989). In dependence on the type of rock this
reasoning is supported by the mercury porosimetry measurements and
furthermore by the analysis of Eq. 1 and Eq. 2 below.
The model by Katz and Thompson (1986, 1987) has its theoretical
foundations in the percolation (or critical path-) concept of
Ambegaokar et al. (1971) and weights both hydraulic and electrical
transport separately through trial solutions (Shante, 1977;
Kirkpatrick, 1979) before connecting the constituent equations to
yield Eq. 1. A reexamination of their model implied that
principally the included shape factor c = 1/226 is in fact not a
constant but a rock dependent adjustable parameter yielding
incorrect results with respect to the experimental data.
The application of Eq. 1 for effective pressures other than zero is
strictly only permissible when the concurrent evolution of the rock
microstructure (respectively the characteristic length scale) is
known. Eq. 2, in contrast, is obtained when both k and F measured
during pressure ramping are plotted against each other on a
logarithmic scale. The experiments then yield sample dependent
parameters r and (c LE
2) as the slope of the linear fit to the data and the intersection
with the y-axis, respectively. The empirical parameter r took
values between 1 and 3 in agreement with previous studies. The
parameters c and LE
2 cannot be derived separately from the experiments but have to be
calculated from the model of choice. LE, in this empirical
approach, is thus a parameter with no true microstructural meaning
and has lost its character as a natural length scale.
This becomes evident when the product (k F) normalized by the
starting values (k0 F0) is plotted as a function of effective
pressure (Figure 5a) and is compared to the pressure dependence of
the product (k Fr) normalized by the starting values (k0 F0
r) (Figure 5b). Any departure of the graphs from a value of 1
(Figure 5a) indicates a pressure dependence of the product (k F)
and thus (c L2) in Eq. 1. In Figure 5b all graphs remain close to 1
as the product (c LE
2) is a constant by definition as outlined above. Here, the
pressure dependence of the microstructure is implicitly contained
within the empirical parameter r. It finally showed that r can vary
with pressure, specifically when the permeability of the sandstone
is dominated by crack porosity. We interpret the empirical
parameter r as a measure of the relative effect of pressure changes
on the linked transport properties. In turn, it can also be viewed
as a qualitative indicator for the degree of coincidence of
hydraulic and electrical flow paths.
This study is documented in detail in Milsch et al. (2008b).
Milsch et al.
6
Figure 5: (a) Normalized product (k F) as a function of effective
pressure. Normalization has been performed with the respective
starting values (k0 and F0) in Table 2. (b) Normalized product (k
Fr) as a function of effective pressure with the parameter r being
sample dependent
4.2 Effect of Dissolution-Precipitation Reactions on Rock
Permeability
4.2.1 Motivation
In the course of reservoir stimulation and exploitation related to
the Groß Schönebeck site the local thermodynamic equilibrium of the
formation is disturbed. This might induce a number of fluid-fluid-
and fluid-rock interactions potentially leading to permeability
damage:
(1) During reservoir stimulation by a water-frac, the formation
fluid is displaced by oxygen-rich well water. This possibly causes
the precipitation of iron hydroxides (Seibt, 2000 and references
cited therein). During the former procedure the temperature in the
proximity of the injection well can be as low as 30°C.
(2) After passing the binary power plant cycle the cooled formation
fluid might become oversaturated with respect to Ba2+ and SO4
2-. Thermodynamic calculations (CHEMEQ) with the typical in situ
concentration range indicated a positive saturation index (SI >
0) for temperatures below 70°C (J. Bartels, personal communication,
2006). As this corresponds to the approximate fluid reinjection
temperature, formation damage due to baryte precipitation is
potentially promoted (Kühn et al, 1997; Dunn et al., 1999).
(3) Finally, any alteration of fluid saturation with respect to the
different ionic species of the rock minerals (e.g. Si) might induce
dissolution-precipitation reactions within the formation in the
course of fluid production (Aharonov et al., 1998; Tenthorey et
al., 1998).
In contrast, an effect of clay swelling on the rock permeability
(e.g. Omar, 1990) in the present case is not to be expected as the
rock does not contain the relevant clay minerals. The preceding
conditions and processes defined the experiments performed in this
study. This study was conducted under realistic conditions
regarding rock- and fluid type, combinations of confining- and pore
pressures, flow-rates as well as temperatures. The principal
physical parameter investigated in the present study was the rock
permeability. The latter was complemented by continuous
measurements of the electrical sample conductivity as well as a
chemical analysis of the pore fluid in regular time
intervals.
4.2.2 Experimental Procedure
The samples tested came from a prospective oil and gas well near
Eberswalde / Germany (Eb 2/76). They were chosen for their
mineralogical similarity with the aquifer rocks at the close by
Groß Schönebeck site where, again, cores were no longer available.
The specimens are Lower Permian (Rotliegend) sandstones of the
Havel subgroup originating from a depth of approximately 4150 m.
The rock is an arcosic litharenite and consists mainly of quartz
(65 vol%). The feldspar content is less than 15 vol%. Rock
fragments are contained by approximately 20 vol% and are mainly of
volcanic origin. Illite and chlorite are the dominant clays. For
the experiments cores were taken parallel to the bedding. Two
specimens labelled ebe05-3 and ebe05-4 were tested.
The fluids used were a 0.1 molar NaCl solution (sample ebe05-3) and
a synthetic Ca-Na-Cl type formation fluid (sample ebe05-4). The
fluid contained 99 g NaCl and 206 g CaCl22H2O per liter distilled
water and thus a total amount of dissolved solids (TDS) of 265
g/L.
The experiments were conducted under hydrostatic conditions in the
HPT-permeameter described above. After assembly, the specimens were
subjected to a confining- and pore pressure of 10 and 5 MPa and a
temperature of 30°C, respectively defining the starting conditions.
Given a full stroke volume of 265 mL the active upstream pore fluid
pump was refilled approximately once per day with the respective
fluid. The flow was unidirectional and repeated pumping of the same
fluid was not performed.
The long-term flow-through experiment with sample ebe05- 4, which
we will focus at in the following, had a total duration of 186
days. The flow-rate was 0.1 mL/min and the test was performed at a
constant confining- and pore pressure of 50 and 5 MPa,
respectively. After confining pressure increase to 50 MPa the
temperature was maintained at 30°C for three days and was then
increased to 150°C. After 30 and 83 days the flow was stopped and
the sample was maintained at stable p-T conditions for 45 and 81
days respectively. After these holds flow was resumed for
approximately 7 days each.
Subsequent to this first experimental stage the fluid was enriched
with progressively increased concentrations of Ba2+ and SO4
2- ions by adding specific amounts of 0.1 molar BaCl2 and Na2SO4
solutions to the synthetic formation fluid. Three different
concentrations n(Ba2+):n(SO4
2-) [n in mM/L] were tested: (1) 0.19:0.19; (2) 0.19:0.57, and (3)
0.38:1.14. This covers the in situ Ba2+ concentrations and
n(Ba2+):n(SO4
2-) ratios of the Groß Schönebeck formation fluid which are in the
range of 0.19 to 0.44 mM/L and 1:1 to 1:3, respectively. This flow-
through test took 7 days and was performed at a temperature of
60°C.
The fluid was then exchanged at 150°C against tap water acidified
to pH5 with acetic acid and the temperature was then, again,
decreased to 60°C and then further to 30°C. The flow with tap water
was maintained for 5 days. At the end of this experimental stage
the tap water was finally exchanged against the original synthetic
formation fluid and the temperature was increased to 150°C to
establish the starting conditions for comparison.
For sample ebe05-4 the total amount of fluid that had transversed
the sample during 60 days of flow was 8.6 L. During both stages of
this experiment a total of 30 fluid samples were taken, generally
one every second day during
Milsch et al.
7
times of flow. The samples contained approximately 150 mL of fluid
each. Immediately after release the fluid pH and redox potential Eh
were measured at ambient p-T conditions (WTW Multi 340i with
Mettler-Toledo probes InLab 412 (pH) and InLab 501 (Eh)).
The fluid samples were then chemically analysed for cation and
anion concentrations of Fe, Mn, Al, Zn, Cu, Pb, K, Si, Ba, and SO4,
respectively. Besides for Fe and Mn which were analysed
photometrically chemical analysis was performed by either GF-AAS
(Graphite Furnance Atomic Absorption Spectrometry; Al, Zn, Cu, Pb)
or by ICP-MS (Inductively Coupled Plasma Mass Spectrometry; K, Si,
Ba, and SO4 through S) at VKTA Rossendorf e.V. Here, the detectable
minimum concentrations, taking dilution into account, were 50 µg/L
(Al, Zn), 10 µg/L (Cu, Pb, K), 60 µg/L (Si), 5 µg/L (Ba), and 1500
µg/L (SO4), respectively.
4.2.3 Results and Discussion
The evolution of permeability during the first stage of this
experiment is shown in Figure 6. The permeability decrease after
start from 4.9 10-15 to 2.0 10-15 m2 is due to a confining pressure
increase from 10 to 50 MPa. During the first 26 days of flow the
permeability remained constant at 2.2 ± 0.8 10-15 m2. The two
longer holds were introduced for comparison by allowing a chemical
equilibration of the fluid with the rock. After flow was resumed
the permeability had decreased by approximately 50 % to 1 ± 0.1
10-15 m2. In contrast, it remained nearly constant for the
remainder of the experiment.
The Figures 7 and 8 show the temperature and permeability,
respectively as a function of time in the course of the second
stage of this experiment. For each of the three different Ba2+ and
SO4
2- concentrations the flow was maintained for two to three days. As
the principal result, neither a change in temperature nor the
different fluid exchanges affected the sample permeability which
remained constant at approximately 1 ± 0.1 10-15 m2.
Figure 6: Sample ebe05-4, stage 1. Permeability as a function of
time
The formation fluid pH before the onset of the experiment was 5.5
at 21.9°C. The recovered fluid samples had pH- values in the range
of 5.1 and 6.6 at 20.0°C with no systematic trend. The redox
potential measurements yielded inconsistent results. Fe and Mn,
both measured photometrically, yielded concentrations of 1.72 ±
0.87 mg/L and 1.76 ± 0.78 mg/L, respectively. The former was the
highest right at the beginning of the test after the temperature
had been increased to 150°C and slightly decreased in the course of
the experiment. For the Mn concentration no systematic trend was
observed. The chemical fluid analysis indicates that ionic species
of K, Cu, Zn and Pb that have been residually preserved
within
the sample after initial coring become easily dissolved and
flushed-out after flow has been started. No dissolution of Al was
observed despite the presence of K-feldspar as well as illite and
chlorite. Si maintained an approximately constant (equilibrium)
concentration of 30 ppm (stage 1) regardless of the flow situation.
The former appears to be affected by the presence of Ba2+ and/or
SO4
2- ions when the latter were introduced artificially (stage 2). An
approximately constant n(Ba2+):n(SO4
2-) ratio of 1:2.2 ± 10 % was observed during stage 1 and thus
prior to the enrichment of the formation fluid with these species.
This indicates the preservation of baryte precipitates after
initial core recovery or the presence of smaller amounts of baryte
cement. Consequently, no effect of precipitation was observed
during stage 2 for a molar ratio of 1:1. Despite a progressive
relative depletion in Ba2+ ions, the SO4
2- concentration measured indicates that baryte precipitation can
also be excluded for the lowest n(Ba2+):n(SO4
2-) ratios investigated in the present study.
Figure 7: Sample ebe05-4, stage 2. Temperature as a function of
time during fluid exchange
Figure 8: Sample ebe05-4, stage 2. Permeability as a function of
time during fluid exchange
Based on the measurements performed it is finally concluded that
during neither long-term experiment with an isochemical fluid
composition any significant alteration of the transport properties
had occurred. Also during stage 2 (sample ebe05-4) no further
change in sample permeability was observed. For baryte
precipitation this is evidently due to an insufficient Ba2+ and
SO4
2- concentration. In addition, precipitation experiments (A. Seibt,
2006, personal communication) indicate that baryte nucleation is
significantly retarded when the fluid is in motion. For the fluid
exchange against tap water this indicates that an effect on the
rock transport properties is only to be expected when the replaced
formation fluid contains a significant amount of ionic species to
be oxidized (e. g. Fe2+).
A detailed documentation of this study can be found in Milsch et
al. (2009).
Milsch et al.
4.3 Temperature Dependence of Electrical Rock Conductivity and
Seismic Wave Velocities
4.3.1 Motivation
Contrary to oil and gas exploration where reflection seismic
surveys are the main exploration method, no similar standard
exploration methods exist for high temperature geothermal
exploration. The on-site conditions are highly variable and the
exploration technology has to be tailor- made for each geothermal
field. In exploring for high temperature fields, reflection seismic
surveys are of limited use due to the structural irregularities.
The most common geophysical methodology involves mainly different
types of resistivity soundings supplemented with other methods like
gravity, magnetic and passive seismic surveys. In order to draw
reliable conclusions about geothermal properties from the results
of geophysical surveys a firm knowledge of the temperature
behaviour of parameters like resistivity and seismic wave velocity
is necessary.
More specifically and in connection with the background given in
Section 3.2, changes in mineralogy were reported to occur at
temperatures close to 230°C in Icelandic geothermal fields
(Kristmannsdóttir, 1979). However, the change in resistivity
appears to be frozen-in via the mineral alteration, i.e. in case
the reservoir has cooled down the boundary between low and high
resistivity persists and does not indicate temperatures exceeding
230°C anymore. It follows that the top of the high resistive core
represents a surface where the temperature once has reached 230°C
but it might have cooled down since. For geothermal exploration it
is important to be able to distinguish between these two cases and
find out if the resistive core has cooled down. One way to move
forward is to learn more about the temperature behaviour of
resistivity, both within the smectite and chlorite alteration zone.
For that purpose, laboratory measurements at in situ conditions on
core samples from various zones of mineral alteration are
important, to determine if there is a way to make any distinction
between reservoirs that have cooled down and those who have
not.
The two main conduction mechanisms in water saturated rocks are
pore fluid conduction and interface conduction. As a first
approximation the two mechanisms act in parallel according to the
(modified) Waxman-Smits equation:
s f
F σ
σ σ +=0 , (3)
where σ0, σf, σs and F are bulk conductivity, pore fluid
conductivity, interface conductivity, and formation factor,
respectively (Rink and Schopper, 1976).
Fluid conductivity, σf, for temperatures below 150°C is linearly
dependent on temperature:
)]()[()( 0f0ff TT1TT −+= ασσ , (4)
where T0 is a reference temperature and αf is a temperature
independent coefficient. The value of αf is assumed to be 0.023/°C
for T0 = 25°C (Revil et al., 1998). Because of the changed
behaviour of density, viscosity and dielectric permittivity of
water for temperatures above 150°C, which affects the mobility of
free charges, the conductivity diverges from linearity above this
temperature limit and normally decreases with increasing
temperature above 250°C (Ucok et al., 1980).
Interface conductivity is dependent on the surface area of pores,
the surface charge density, the valence and mobility of surface
ions, temperature, and acidity (pH). For temperatures below 200°C,
however, also interface conductivity, σs, can be described by a
linear relationship with temperature:
)]()[()( 0s0ss TT1TT −+= ασσ , (5)
where, as before, T0 is a reference temperature and αs is a
temperature independent coefficient. Revil et al. (1998) determined
a value of αs = 0.040/°C for T0 = 25°C. In general, surface
conductivity is believed to be more temperature dependent than pore
fluid conductivity (Revil et al., 1998).
There is a clear difference in conductivity between rocks in the
smectite or chlorite alteration zone, according to field
measurements. The cation exchange capacity (CEC), which is the
quantity of exchangeable cations on a negatively charged mineral
surface, is not the same for these two clay minerals: For smectite
the CEC is 0.8-1.5 meq/g (Ellis, 1987) and for chlorite it is 0.01
meq/g (Thomas, 1976), possibly explaining the great difference in
conductivity between the two mineral alteration zones.
4.3.2 Experimental Procedure
All measurements were made at the GFZ in the HPT- permeameter
described above. The measured samples all originated from Icelandic
boreholes. All samples were kept dry at room conditions since their
drilling, except sample K40 which was kept submerged in fluid. An
overview of the samples used in this investigation can be found in
Table 3. With exception of sample K40, the fluids of all other
specimens were synthetically prepared. This was achieved by
dissolving specific amounts of reagent grade NaCl, KCl, Na2SO4, and
K2SO4 salts in distilled water. The specific concentrations chosen
were based on fluid analyses of samples taken at the respective
geothermal site and reflect the principal in situ chemical
compositions. A total of 7 samples were heated in steps of 25°C or
50°C in the temperature range of 25°C to 250°C, keeping confining
and pore pressure constant at the respective in situ condition.
Meanwhile, the temperature at the sample, pore and confining
pressure, and electrical conductivity of the sample were recorded
continuously. When equilibrium in temperature and conductivity had
been reached for each step both impedance in the frequency range
0.01 Hz to 100 kHz and P-wave velocity were measured in
addition.
Table 3. Sample properties (Sections 4.3, 4.4, 4.5), MLC: mixed
layer clays.
Milsch et al.
4.3.3 Results and Discussion
The general behavior of conductivity with respect to temperature
was similar for all samples tested. As an example we present the
data for one particular sample (2B). Temperature and conductivity
as a function of time are depicted in Figure 9. Conductivity as a
function of temperature is shown in Figure 10. Note the apparent
transient conductivity increase at 150°C, which indicates changes
in the intrinsic sample properties (Section 4.4).
Figure 9: Conductivity and temperature as a function of time.
Example: sample 2B.
Figure 10: Conductivity as a function of temperature, showing
measurements made both in continuous scans and with an LCR-meter.
Example: sample 2B
The linear relation between conductivity and temperature in the
range 50-170°C is evident, both before and after the change in
conductivity. At temperatures above 170-200°C the conductivity
becomes less and less temperature dependent. If both conduction
mechanisms contribute, the slope α of the curve σ(T)/σ(T0) takes on
a value somewhere
between αf and αs, i.e. )(
)()(
= .
For T0 = 25°C we obtain values in the range α = 0.027-0.28 /°C,
i.e. all higher than αf = 0.023 /°C. This could indicate that pore
fluid conduction is dominated by interface conduction both in the
smectite zone and the chlorite zone, although samples 2B and 3A
from well ÖJ-1 are most likely a mixture of interface and pore
fluid conduction after the conductivity change at 150°C (α =
(0.027±0.003) /°C). We conclude that there is no definite
difference in conduction mechanisms between the smectite zone and
the mixed layer clays / chlorite zone. Therefore we assume that the
much
lower CEC of chlorite compared to smectite is the most likely cause
of the clear conductivity decrease between the smectite and
chlorite zones in geothermal HT-fields.
Finally, the measurements of P-wave velocities indicate a
systematic decrease by 5 to 15 % with temperature for all samples.
Before the observed transient change in conductivity the velocities
were slightly higher than afterwards at the respective temperature.
P-wave velocities were in the range of 4.4 km/s (25°C) and 3.4 km/s
(250°C).
A complete documentation of this study can be found in
Kristinsdóttir et al. (in review). The ultrasonic data is
documented and analysed in detail in Jaya et al. (in review).
In an ongoing study, similar experiments on conductivity and
ultrasonic wave velocities are conducted on Italian samples from
the Anqua and Radicondoli wells in the Travale geothermal test
site.
4.4 Evolution of Electrical Rock Conductivity in a Fluid- Rock
Disequilibrium
4.4.1 Motivation
As indicated in Section 4.3 electrical conductivity showed a
transient behavior when the sample was first heated to 150°C. This
implies a change in the samples’ intrinsic properties as a result
of a fluid-rock disequilibrium until a thermodynamically stable
state is reached with respect to the actual p-T-X conditions. An
investigation of this effect is of particular importance as
reservoir states of equilibrium are always disturbed during
stimulation procedures and as fluid is produced. As a consequence,
results of resistivity logging will be time dependent as well and
the obtained data has to be interpreted accordingly. Also, so
measured approaches to equilibrium could potentially be interpreted
as saturation states of the pore fluid. A disturbance of fluid-
rock equilibria, e.g. by cooling during energy production, might
result in mineral precipitation and consequently in permeability
damage in the near-bore region of the injection well or even deeper
within the formation.
4.4.2 Experimental Procedure
During the experiments described in Section 4.3 two of the samples
(2B and K40) were flushed with fresh pore fluid to measure
permeability. Measurements were made at 40°C before the temperature
was raised for the first time and then repeatedly during each
temperature cycle at 150°C, after conductivity had stabilized. The
flow rate was Q = 25 µL/min for sample K40 and Q = 50 µL/min for
sample 2B.
4.4.3 Results and Discussion
The permeability decreased considerably as the samples were heated
from 40°C to 150°C, i.e. during the conductivity hysteresis: from
9.3 to 5.5 µD (K40) and 25.4 to 21.8 µD (2B). The reason for this
decrease, at present, remains unresolved. At constant temperature
(150°C) the permeability of both samples remained approximately
constant within experimental resolution, i.e. k = (5.3±0.2) µD for
sample K40, and k = (18.7±0.6) µD for sample 2B.
During every permeability test the conductivity decreased at 5-10
µS/cm·h for sample K40 and 8-12 µS/cm·h for sample 2B, except in
the last measurement (2B) when the decrease was as high as 72
µS/cm·h. This last conductivity decrease in sample 2B is shown in
red on Figure 11. The final conductivity value after this decrease
replicated exactly the one before the transition at 150°C.
Afterwards,
Milsch et al.
10
the temperature was decreased and increased again to measure the
actual σ-T relation.
The transient conductivity behaviour around 150°C and its
reversibility have yet to be explained. However, the temperature
coefficient α decreased for every sample during the transition
which indicates that the conductance is more associated with the
fluid conductivity than before. Also, the fluid in the system
before the measurements had a conductivity of σ = 970 µS/cm. After
the measurements the conductivity of the fluid in the downstream
pump was σ = 1280 µS/cm.
Figure 11: Conductivity as a function of temperature (continuous
scans), with each heating cycle shown in different color. Example:
sample 2B
Based on these observations it is implied that the transient
conductivity behavior is due to some flow of ions over the boundary
between the (clay) minerals or the Stern layer and the pore fluid,
more precisely, that there is some uncompensated exchange of ions
or mineral dissolution. The result is that a larger proportion of
the ions that contribute most to the conductivity are in contact
with free fluid. Hence, when fresh fluid flows through the sample
these additional ions are washed out and the sample conductivity
decreases.
The data related to this section is documented in more detail in
Kristinsdóttir et al. (in review).
In an ongoing study, the former effects are investigated for rocks
originating from or related to sedimentary geothermal reservoirs
(Schepers and Milsch, 2009).
4.5 Petrophysical Signature of a Water-Steam Phase Transition
within the Pore Space
4.5.1 Motivation
During production from high temperature geothermal reservoirs the
pore pressure decreases in response to the mass withdrawal. This
results in the onset of boiling often followed by formation of a
steam cap above the production zone. The steam cap can both be a
target for energy production as well as a source of potential
hazards if the pressure in the steam cap exceeds the weight of the
overburden. For both reasons it is therefore of particular
importance to develop methods to detect and map the extent of such
steam caps. As the electrical conductivity of steam is much lower
than that of liquid water electrical exploration methods might be
particularly suitable if the conductivity contrast at the field
scale is high enough.
The overall effect of boiling on electrical conductivity within
geothermal reservoirs is poorly constrained although it can be
expected that conductivity will decrease if boiling starts. In
contrast to this likely scenario for fluid dominated electrical
conductivity, the direct effect of vaporization on surface
conduction in the pores is, so far, unknown.
Based on the Young-Laplace concept of capillarity (e.g. Bear, 1988)
Roberts et al. (2001a) proposed a model for a finely porous medium,
where pore fluid vaporization would be heterogeneous as pore
pressure is decreased, because of the effect of capillary suction.
For a finely porous medium, vaporization proceeds when the pore
pressure is:
capboilpore ppp +≤ , (6)
where ppore, pboil, and pcap are the pore pressure, boiling (vapor)
pressure, and capillary pressure, respectively. At least for fluid
dominated conduction, the signature of vaporization with respect to
electrical conductivity will therefore be dependent on the
individual rock microstructure.
4.5.2 Experimental Procedure
The experiments were conducted at the GFZ in the HPT- permeameter
described above and subsequent to the investigations reported in
Section 4.3 and 4.4. Again, the measurements were carried out on
the four volcanic rock samples from Iceland, two basalts (58 and
K40) and two hyaloclastites (2B and 3A). One sample of sandstone
from Fontainebleau, France (FTBS12) was added to the experimental
program as a reference (Table 3). The fluid used for the
Fontainebleau sandstone was a 0.1 molar NaCl synthetic solution
having an electrical conductivity of 10.8 mS/cm at 25°C.
During the experiments both temperature (nominally 150°C) and
confining pressure were kept constant. In contrast, pore pressure
was decreased to allow vaporization of pore fluid in a controlled
manner. The downstream pore fluid pump was retracted at a constant
rate so that the total volume of the pore fluid system was steadily
increased. As both sides of the sample were connected during this
procedure, the nominal pore pressure was equal at either face of a
specimen. The pore fluid pumps were kept at room temperature and
vaporization was restricted to the hot zone of the pore fluid
system located within the pressure vessel. During the experiments
the electrical conductivity, the pore pressure, the volume of the
pore fluid pump and the sample temperature were continuously
monitored.
4.5.3 Results and Discussion
For all samples, a continuous conductivity decrease was observed as
the volume of the pore fluid system was increased progressively.
The total conductivity decrease during vaporization was generally 1
to 2 orders of magnitude and thus significant. Besides for sample
58 where the measurement was terminated earlier, the electrical
conductivity ultimately reached a minimum.
In contrast but also for all samples, there was a discontinuous
decrease in pore pressure. Initially, as the volume of the pore
fluid system is increased, the pressure dropped rapidly due to
elastic relaxation. Then, at the boiling point, an approximately
constant pore pressure level was maintained. Finally, the pore
pressure continuously decreased again.
Milsch et al.
11
The relationship between the evolution of both electrical
conductivity and pore (vapor) pressure was observed to be sample
dependent. If capillarity is the only reason, differences in pore
pressure evolution then would reflect differences in sample
microstructure in terms of the individual pore radii distributions.
Following the procedure in Roberts et al. (2001a) one can calculate
the minimum pore (capillary) radius R related to vaporization by
the maximum capillary pressure observed (pcap ≈ (-) 0.25 MPa at
150°C):
Rpcap /cos2 θγ−= , (7)
where γ and θ are surface tension of the wetting fluid (water; 5.2
10-3 Pa m; Weast, 1984) and wetting angle (≈ 0; Roberts et al.,
2001a), respectively. One obtains R ≥ 42 nm, which indicates that
virtually all pore size classes were affected by vaporization in
the present experiments.
An important conclusion that can be drawn is the amount of vapor
required to reduce the electrical conductivity to a minimum with
respect to the individual sample porosity. In Figure 12 one notices
that for all samples 5 ± 1 mL of fluid have to be drained before
the electrical sample conductivity becomes affected. This volume
relates directly to the free fluid volume (e.g. in tubings) located
at 150°C inside the pressure vessel. For all but one sample (58)
the behavior, then, is similar. The electrical conductivity
decreases to only about 5 ± 1 % of its starting value. Furthermore
(excluding 58), 7 ± 2 mL of fluid have to be drained from the
sample until the minimum electrical conductivity is reached. The
total (connected) pore volume of the samples is 3.5 (FTBS12), 3.7
(K40), 4.1 (2B), 5.7 (58), and 5.9 (3A) mL. The ratio between the
drained fluid volume and the total pore volume is 1.5 (FTBS12), 2.5
(K40), 1.3 (2B), 3.1 (58), and 1.0 (3A), thus approximately 2.0 ±
1.0 times the relevant pore volume. Due to fluid condensation
outside the vessel the drained fluid volume in fact directly
reflects the transformed liquid volume contained within the pore
fluid system and thus the pore space.
This balance then allows a classification of the rocks. Samples 3A,
2B, and FTBS12 display a very similar conductivity signature. Here,
every part of drained volume larger than the pore volume can well
be attributed to smaller pores and/or a broader pore size
distribution. The excess volume then reflects the vapor expansion
related to a pressure decrease necessary for further vaporization.
For sample 58, in contrast, the excess volume at the conductivity
minimum was comparatively large. This suggests that the curve
shapes in Figure 12 could also be indicative of the respective
dominating conduction mechanism as vaporization proceeds.
Electrical conduction in sample FTBS12 is definitely fluid
dominated for the given fluid salinity (Milsch et al., 2008b). This
applied to samples 3A and 2B indicates that conduction during
vaporization in hyaloclastites emanating from the chlorite
alteration zone should be fluid dominated as well, even for pore
fluids of low salinity as in the present case. Sample K40, a basalt
containing mixed-layer clays, then would display a transition from
fluid to surface conduction at a later vaporization stage. Finally
and for the present fluid composition, conduction in sample 58, a
basalt from the smectite alteration zone, is supposed to be surface
dominated. In this case, the decrease in electrical conductivity
upon continuous fluid drainage should be related to a progressive
destruction of the conductive layer on the mineral surfaces rather
than to the phase transformation within the pore space itself. The
concurrent
evolution of electrical conductivity and pore pressure, in this
case, would be largely unrelated. This finally suggests that
variances in the respective conduction mechanism yield different
electrical signatures as vaporization proceeds.
Finally, at the observed conductivity minimum it can be assumed
that apart from retained water on grain surfaces or in ultra-small
pores all samples can be considered dry. Therefore, samples with
fluid dominated electrical conduction are suggested to
approximately quantify their respective liquid-vapor saturation via
the measured electrical conductivity signature. The exact phase
distribution during vaporization, however, remains
unresolved.
Figure 12: Normalized electrical conductivity as a function of
volume change in the pore pressure system. The conductivity
normalization was performed with the respective starting value
right before the initiation of boiling within the pore space
A detailed documentation of this study can be found in Milsch et
al. (in review).
CONCLUSIONS
Two identical flow-through apparatuses have been set-up to simulate
p-T-X conditions related to deep geothermal reservoirs. Their
long-term performance was successfully tested for confining
pressures, pore pressures and temperatures of 100 MPa, 50 MPa and
200 °C, respectively, thus exploring stress-isotropic upper crustal
in situ conditions at up to 5 km depth including the respective
temperatures for a normal geothermal gradient. Highly saline pore
fluids can be used. In addition, four important rock physical
parameters (permeability, electrical conductivity as well as P and
S wave velocities) can be determined simultaneously. Continuous
flow experiments, so far, have been performed over a maximum of six
months. Stable physical conditions have thus been maintained over
periods that are significantly longer than those usually attained
in conventional laboratory based rock physical testing.
Scientifically, the usage of the devices is focused on the
evaluation of risk potentials in exploration and exploitation of
deep geothermal reservoirs regarding sustainable production.
Particularly, the investigations presented above addressed possible
effects of fluid-rock interactions on the transport properties of a
reservoir host rock. Also, the pressure and temperature dependence
of rock properties as well as effects of phase changes within the
pore space were characterized in several independent studies. The
investigations were conducted both site specific and process
Milsch et al.
12
oriented. We presented evidence that laboratory based research is a
valuable complement to both field and numerical studies on the
long-term evolution of geothermal reservoirs.
ACKNOWLEDGEMENTS
Financial support was provided by the German Federal Ministry for
the Environment, Nature Conservation, and Nuclear Safety under
grant BMU 0329951B and the European Commission under the 6th
Framework Programme of the European Union.
REFERENCES
Aharonov, E., Tenthorey, E., and Scholz, C. H.: Precipitation
sealing and diagenesis: 2. Theoretical analysis. J. Geophys. Res.
103 (B10), 23969-23981 (1998).
Ambegaokar, V., Halperin, B.I., and Langer, J.S.: Hopping
conductivity in disordered systems. Phys. Rev. B 4, 2612-2620
(1971).
Archie, G.E.: The electrical resistivity log as aid in determining
some reservoir characteristics. Trans. Am. Inst. Mech. Eng. 146,
54-61 (1942).
Bear, J.: Dynamics of fluids in porous media. Dover Publ., Inc.
Mineola, NY (1988).
Blöcher, G., Moeck, I., Milsch, H., and Zimmermann, G.: Modelling
of pore pressure response due to hydraulic stimulation treatments
at the geothermal research doublet EGrSk3/90 and GtGrSk4/05 in
summer 2007. Proceedings, 33rd Workshop on Geothermal Reservoir
Engineering, Stanford University, Stanford, California, SGP-TR-185
(2008).
Blöcher, G., Zimmermann, G., and Milsch, H.: Impact of poroelastic
response of sandstones on geothermal power production. Pure appl.
geophys. 166, 1-17, doi 10.1007/s00024-009-0475-4 (2009).
Brace, W.F.: Permeability from resistivity and pore shape. J.
Geophys. Res. 82 (23), 3343-3349 (1977).
Civan, F.: Reservoir formation damage – Fundamentals, Modelling,
Assessment, and Mitigation. 1st Ed., Gulf Publ. Co., Houston, TX,
and Butterworth-Heinemann, Woburn, MA, 742 p. (2000).
Darcy, H.: Les fontaines publique de la ville de Dijon. Dalmont,
Paris (1856).
David, C.: Geometry of flow paths for fluid transport in rocks. J.
Geophys. Res. 98, 12267-12278 (1993).
Dunn, K., Daniel, E., Shuler, P. J., Chen, H. J., Tang, Y., and
Yen, T. F.: Mechanisms of surface precipitation and dissolution of
baryte: A morphology approach. J. Colloid Interf. Sci. 214 (2),
427-437 (1999).
Ellis, D.V.: Well Logging for Earth Scientist. Elsevier, New York
(1987).
Flovenz, O.G., Georgsson, L.S., and Arnason, K.: Resistivity
structure of the Upper Crust in Iceland. J. Geophys. Res. 90(B12),
10136-10150 (1985).
Guéguen, Y. and Dienes, J.: Transport Properties of Rocks from
Statistics and Percolation. Math. Geol. 21 (1), 1- 13 (1989).
Jaeger, J.C. and Cook, N.G.W.: Fundamentals of rock mechanics.
Science Paperbacks. Chapman and Hall, London, 2nd Edition
(1976).
Jaya, M.S., Shapiro, S.A., Kristinsdóttir, L.H., Bruhn, D., Milsch,
H., and Spangenberg, E.: Temperature- dependence of seismic
properties in geothermal rocks at simulated reservoir conditions.
Geothermics (in review).
Katz, A.J. and Thompson, A.H.: Quantitative prediction of
permeability in porous rock. Phys. Rev. B 34 (11), 8179-8181
(1986).
Katz, A.J. and Thompson, A.H.: Prediction of Rock Electrical
Conductivity From Mercury Injection Measurements. J. Geophys. Res.
92 (B1), 599-607 (1987).
Kirkpatrick, S.: Models of disordered materials, in: Balian, R.,
Maynard, R., Toulouse, G. (Eds.), Ill-Condensed Matter.
North-Holland, Amsterdam, 323-403 (1979).
Kristinsdóttir, L.H., Flovenz, O.G., Arnason, K., Bruhn, D.,
Milsch, H., Spangenberg, E., and Kulenkampff, J.: Laboratory
measurements of conductivity of rock samples from Icelandic high
temperature geothermal fields as a function of temperature at
in-situ conditions. Geothermics (in review).
Kristmannsdóttir, H.: Alteration of basaltic rocks by hydrothermal
activity at 100-300°C. In: Developments in Sedimentology, 27 (eds.
Mortland, M., and Farmer, V.), pp. 359-367. Elsevier, Amsterdam
(1979).
Kühn, M., Frosch, G., Kölling, M., Kellner, T., Althaus, E., and
Schulz, H. D.: Experimentelle Untersuchungen zur Barytübersättigung
einer Thermalsole, Grundwasser 3, 111–117 (1997).
Kulenkampff, J., Spangenberg, E., Flovenz, O.G., Raab, S., and
Huenges, E.: Petrophysical Parameters of Rocks Saturated with
Liquid Water at High Temperature Geothermal Reservoir Conditions.
In: Proceedings World Geothermal Congress. Antalya, Turkey
(2005).
Legarth, B., Huenges, E., and Zimmermann, G.: Hydraulic fracturing
in a sedimentary geothermal reservoir: Results and implications.
Int. J. Rock Mech. Min. Sci. 42, 1028–1041 (2005).
Martys, N. and Garboczi, E.J.: Length scales relating the fluid
permeability and electrical conductivity in random two-dimensional
model porous media. Phys. Rev. B 46 (10), 6080-6090 (1992).
Milsch, H., Spangenberg, E., Kulenkampff, J., and Meyhöfer, S.: A
new apparatus for long-term petrophysical investigations on
geothermal reservoir rocks at simulated in-situ conditions. Transp.
Porous Med. 74, 73-85, doi 10.1007/s11242-007-9186-4 (2008a).
Milsch, H., Blöcher, G., and Engelmann, S.: The relationship
between hydraulic and electrical transport properties in
sandstones: An experimental evaluation of several scaling models.
Earth Planet. Sci. Lett. 275, 355-363, doi
10.1016/j.epsl.2008.08.031 (2008b).
Milsch, H., Seibt, A., and Spangenberg, E.: Long-term Petrophysical
Investigations on Geothermal Reservoir Rocks at Simulated In Situ
Conditions. Transp. Porous Med. 77, 59-78, doi
10.1007/s11242-008-9261-5 (2009).
Milsch, H., Kristinsdóttir, L.H., Spangenberg, E., Bruhn, D., and
Flovenz, O.G.: Effect of the water steam phase transition in porous
rocks on their electrical conductivity. Geothermics (in
review).
Milsch et al.
13
Moeck, I., Schandelmeier, H., and Holl, H.-G.: The stress regime in
a Rotliegend reservoir of the Northeast German Basin. Int. J. Earth
Sci., doi 10.1007/s00531- 008-0316-1 (2008).
Omar, A. E.: Effect of brine composition and clay content on the
permeability damage of sandstone cores. J. Petroleum Sci. Eng. 4,
245-256 (1990).
Paterson, M.S.: The equivalent channel model for permeability and
resistivity in fluid saturated rocks – A reappraisal. Mech. Mater.
2 (4), 345-352 (1983).
Piwinskii, A.J. and Weed, H.C.: A study of rock-solution
interaction and its effect on Archie’s Law. IEEE Trans. Geosci.
Electr. GE-14 (4), 221-223 (1976).
Revil, A., Cathles III, L. M., Losh, S., and Nunn, J. A.:
Electrical conductivity in shaly sands with geophysical
applications. J. Geophys. Res. 103 (B10), 23925- 23936
(1998).
Rink, M. and Schopper, J.R.: Pore structure and physical properties
in porous sedimentary rocks. Pure Applied Geophysics 114, 273-284
(1976).
Roberts, J.J., Duba, A.G., Bonner, B.P., and Kasameyer, P.W.: The
effects of capillarity on electrical resistivity during boiling in
metashale from scientific corehole SB-15-D, The Geysers,
California, USA. Geothermics 30, 235-254 (2001).
Ruffet, C., Darot, M., and Guéguen, Y.: Surface conductivity in
rocks: A review. Surveys in Geophys. 16, 83-105 (1995).
Scheidegger, A. E.: The physics of flow through porous media. Univ.
of Toronto Press, Toronto (1974).
Schepers, A. and Milsch, H.: Effects of fluid-rock interactions in
arkosic sandstones: Long-term direct monitoring of changes in
permeability, electrical conductivity, and pore fluid chemistry.
Geophys. Res. Abstracts 11, EGU2009-5767 (2009).
Seibt, A.: Welche Faktoren können die Eisen(II)-Oxidation in
Formationswässern beeinflussen? In: Huenges, E. (ed.) Geothermische
Energieentwicklung–geologische
und energietechnische Ansatzpunkte. Scientific Technical Report,
STR00/23, GeoForschungsZentrum Potsdam, Potsdam, Germany
(2000).
Shante, V.K.S.: Hopping conduction in quasi-one- dimensional
disordered compounds. Phys. Rev. B 16 (6), 2597-2612 (1977).
Tenthorey, E., Scholz, C. H., Aharonov, E., and Léger, A.:
Precipitation sealing and diagenesis: 1. Experimental results. J.
Geophys. Res. 103 (B10), 23951-23967 (1998).
Thomas, E.C.: Determination of Qv from membrane potential
measurements on shaly sands. Transactions of the American Institute
of Mining, Metallurgical, and Petroleum Engineers 261, 1087-1096
(1976).
Ucok, H., Ershaghi, I., and Olhoeft, G.R.: Electrical Resistivity
of Geothermal Brines. Journal of Petroleum Technology, pp. 717-727
(1980).
Van Siclen, C.D.: Equivalent channel network model for permeability
and electrical conductivity of fracture networks. J. Geophys. Res.
107 (B6), 2106, doi:10.1029/2000JB000057 (2002).
Walsh, J.B. and Brace, W.F.: The Effect of Pressure on Porosity and
the Transport Properties of Rock. J. Geophys. Res. 89 (B11),
9425-9431 (1984).
Weast, R.C. (Ed.): CRC Handbook of Chemistry and Physics, 64th
Edition. CRC Press, Inc., Boca Raton, FL, USA (1984).
Wyllie, M.R.J. and Rose, W.D.: Some theoretical considerations
related to the quantitative evaluation of the physical
characteristics of reservoir rock from electrical log data. Trans.
Am. Inst. Mech. Eng. 189, 105-118 (1950).
Zimmermann, G., Reinicke, A., Brandt, W., Blöcher, G., Milsch, H.,
Holl, H.-G., Moeck, I., Schulte, T., Saadat, A., and Huenges, E.:
Results of stimulation treatments at the geothermal research wells
in Groß Schönebeck/Germany. In: Proceedings 33rd Stanford Workshop
on Geothermal Reservoir Engineering. Stanford, USA (2008).
<< /ASCII85EncodePages false /AllowTransparency false
/AutoPositionEPSFiles true /AutoRotatePages /All /Binding /Left
/CalGrayProfile (Dot Gain 20%) /CalRGBProfile (sRGB IEC61966-2.1)
/CalCMYKProfile (U.S. Web Coated \050SWOP\051 v2) /sRGBProfile
(sRGB IEC61966-2.1) /CannotEmbedFontPolicy /Warning
/CompatibilityLevel 1.5 /CompressObjects /Tags /CompressPages true
/ConvertImagesToIndexed true /PassThroughJPEGImages true
/CreateJDFFile false /CreateJobTicket false /DefaultRenderingIntent
/Default /DetectBlends true /ColorConversionStrategy
/LeaveColorUnchanged /DoThumbnails false /EmbedAllFonts true
/EmbedJobOptions true /DSCReportingLevel 0 /SyntheticBoldness 1.00
/EmitDSCWarnings false /EndPage -1 /ImageMemory 1048576
/LockDistillerParams false /MaxSubsetPct 100 /Optimize true /OPM 1
/ParseDSCComments true /ParseDSCCommentsForDocInfo true
/PreserveCopyPage true /PreserveEPSInfo true /PreserveHalftoneInfo
false /PreserveOPIComments false /PreserveOverprintSettings true
/StartPage 1 /SubsetFonts true /TransferFunctionInfo /Apply
/UCRandBGInfo /Preserve /UsePrologue false /ColorSettingsFile ()
/AlwaysEmbed [ true ] /NeverEmbed [ true ] /AntiAliasColorImages
false /DownsampleColorImages false /ColorImageDownsampleType
/Bicubic /ColorImageResolution 300 /ColorImageDepth -1
/ColorImageDownsampleThreshold 1.50000 /EncodeColorImages true
/ColorImageFilter /DCTEncode /AutoFilterColorImages true
/ColorImageAutoFilterStrategy /JPEG /ColorACSImageDict <<
/QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >>
/ColorImageDict << /QFactor 0.15 /HSamples [1 1 1 1]
/VSamples [1 1 1 1] >> /JPEG2000ColorACSImageDict <<
/TileWidth 256 /TileHeight 256 /Quality 30 >>
/JPEG2000ColorImageDict << /TileWidth 256 /TileHeight 256
/Quality 30 >> /AntiAliasGrayImages false
/DownsampleGrayImages false /GrayImageDownsampleType /Bicubic
/GrayImageResolution 300 /GrayImageDepth -1
/GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true
/GrayImageFilter /DCTEncode /AutoFilterGrayImages true
/GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict <<
/QFactor 0.15 /HSamples [1 1 1 1] /VSamples [1 1 1 1] >>
/GrayImageDict << /QFactor 0.15 /HSamples [1 1 1 1] /VSamples
[1 1 1 1] >> /JPEG2000GrayACSImageDict << /TileWidth
256 /TileHeight 256 /Quality 30 >> /JPEG2000GrayImageDict
<< /TileWidth 256 /TileHeight 256 /Quality 30 >>
/AntiAliasMonoImages false /DownsampleMonoImages true
/MonoImageDownsampleType /Bicubic /MonoImageResolution 1200
/MonoImageDepth -1 /MonoImageDownsampleThreshold 1.50000
/EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode
/MonoImageDict << /K -1 >> /AllowPSXObjects false
/PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false
/PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000
0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true
/PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ]
/PDFXOutputIntentProfile () /PDFXOutputCondition ()
/PDFXRegistryName (http://www.color.org) /PDFXTrapped /Unknown
/Description << /FRA
<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>
/JPN
<FEFF3053306e8a2d5b9a306f30019ad889e350cf5ea6753b50cf3092542b308000200050004400460020658766f830924f5c62103059308b3068304d306b4f7f75283057307e30593002537052376642306e753b8cea3092670059279650306b4fdd306430533068304c3067304d307e305930023053306e8a2d5b9a30674f5c62103057305f00200050004400460020658766f8306f0020004100630072006f0062006100740020304a30883073002000520065006100640065007200200035002e003000204ee5964d30678868793a3067304d307e30593002>
/DEU
<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>
/PTB
<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>
/DAN
<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>
/NLD
<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>
/ESP
<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>
/SUO
<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>
/ITA
<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>
/NOR
<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>
/SVE
<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>
/ENU
<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>
>> >> setdistillerparams << /HWResolution [2400
2400] /PageSize [612.000 792.000] >> setpagedevice